Prima Mente needed to model genome-scale epigenetic and DNA signals for neuroscience and precision diagnostics, but whole-genome methylation data, clinical benchmark sets, and foundation-model training all exceeded what conventional biology tooling could process quickly enough for rapid iteration.
Reported impact
Pleiades generated DNA fragments with 83% per-nucleotide accuracy, increased the amount of DNA from specific cell types in a blood sample by up to 20x, and reached 0.89 accuracy for early-stage Alzheimer's and Parkinson's detection, rising to 0.97 when protein tests were added.
Prima Mente built the Pleiades family of biological foundation models on NVIDIA DGX Cloud running in Nebius AI Cloud to jointly learn DNA sequence and methylation patterns across human cell types. The team curated large human methylation and genomic datasets, then pretrained and fine-tuned autoregressive transformer decoders in 90M, 600M, and 7B parameter variants on 1.9 trillion tokens of epigenetic data. NVIDIA Quantum InfiniBand networking connected the GPU cluster so large training jobs and evaluation runs could move efficiently across the environment. The resulting platform lets Prima Mente reconstruct cell-free DNA fragments, enrich rare cell-type signals, and score early-stage neurological disease risk from blood-based data, giving researchers a reusable compute and model stack for precision-diagnostics and neuroscience workflows.
Human methylation and genomic datasets are curated into training corpora, loaded into Pleiades jobs on DGX Cloud in Nebius AI Cloud, processed by transformer models over the InfiniBand-connected cluster, and then used for fragment reconstruction and disease-prediction inference for diagnostics research.
For early detection of Alzheimer's and Parkinson's, Pleiades achieved 0.89 accuracy, rising to 0.97 with protein tests.
Pleiades generated DNA fragments with 83% accuracy using blood samples.
Pleiades has been trained using 1.9 trillion tokens of epigenetic data.
Their first goal is to understand the brain across health, age and disease to cure severe human diseases like Alzheimer's.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
The summary text states 'Prima Mente employs 43 employees.'
Related
NVIDIA Sample Sep 22, 2026 DELFI Diagnostics DELFI Diagnostics needed to speed genomic analysis for cancer screening and therapeutic monitoring, but CPU-bound alignment and somatic variant-calling steps created turnaround-time and cost bottlenecks that limited how… 10–20 business days DeepSomatic · MutectCaller · NVIDIA Parabricks NVIDIA Sample Jul 10, 2026 Astellas Pharma Inc. Astellas needed to shorten the slow handoff between experimental scientists and computational chemists in early drug discovery so more researchers could use structure prediction and affinity screening without waiting on… More than 50 regular users and over 2,000 total users within four months NVIDIA BioNeMo · NVIDIA Boltz-2 NIM · NVIDIA DGX H100/H200 NVIDIA Sample Apr 23, 2026 Biofy Hospitals needed pathogen identification and antibiotic-resistance analysis much faster than five-day culture and antibiogram workflows, but CPU-bound genomic processing and limited AI expertise made it hard to turn… 100,000+ synthetic DNA sequences generated by the biology model Abby · H100 Tensor Core GPUs · NVIDIA CUDA-accelerated libraries 